Nucleic Acids Res. 2024 Oct 26:gkae966. doi: 10.1093/nar/gkae966. Online ahead of print.

ABSTRACT

Invertible promoters (invertons) are crucial regulatory elements in bacteria, facilitating gene expression changes under stress. Despite their importance, their prevalence and the range of regulated gene functions are largely unknown. We introduced DeepInverton, a deep learning model that identifies invertons across a broad phylogenetic spectrum without using sequencing reads. By analyzing 68 733 bacterial genomes and 9382 metagenomes, we have uncovered over 200 000 nonredundant invertons and have also highlighted their abundance in pathogens. Additionally, we identified a post-Cambrian Explosion increase of invertons, paralleling species diversification. Furthermore, we revealed that invertons regulate diverse functions, including antimicrobial resistance and biofilm formation, underscoring their role in environmental adaptation. Notably, the majority of inverton identifications by DeepInverton have been confirmed by the in vitro experiments. The comprehensive inverton profiles have deepened our understanding of invertons at pan-genome and pan-metagenome scales, enabling a broad spectrum of applications in microbial ecology and synthetic biology.

PMID:39460615 | DOI:10.1093/nar/gkae966

Br J Psychol. 2024 Oct 25. doi: 10.1111/bjop.12744. Online ahead of print.

ABSTRACT

This study explored the facial impressions of attractiveness, dominance and sexual dimorphism using experimental and computational methods. In Study 1, we generated face images with manipulated morphological features using geometric morphometrics. In Study 2, we conducted eye tracking and impression evaluation experiments using these images to examine how facial features influence impression evaluations and explored differences based on the sex of the face images and participants. In Study 3, we employed deep learning methods, specifically using gradient-weighted class activation mapping (Grad-CAM), an explainable artificial intelligence (AI) technique, to extract important features for each impression using the face images and impression evaluation results from Studies 1 and 2. The findings revealed that eye-tracking and deep learning use different features as cues. In the eye-tracking experiments, attention was focused on features such as the eyes, nose and mouth, whereas the deep learning analysis highlighted broader features, including eyebrows and superciliary arches. The computational approach using explainable AI suggests that the determinants of facial impressions can be extracted independently of visual attention.

PMID:39460393 | DOI:10.1111/bjop.12744

Sensors (Basel). 2024 Oct 21;24(20):6767. doi: 10.3390/s24206767.

ABSTRACT

In this paper, a deep-learning-based frame synchronization blind recognition algorithm is proposed to improve the detection performance in non-cooperative communication systems. Current methods face challenges in accurately detecting frames under high bit error rates (BER). Our approach begins with flat-top interpolation of binary data and converting it into a series of grayscale images, enabling the application of image processing techniques. By incorporating a scaling factor, we generate RGB images. Based on the matching radius, frame length, and frame synchronization code, RGB images with distinct stripe features are classified as positive samples for each category, while the remaining images are classified as negative samples. Finally, the neural network is trained on these sets to classify test data effectively. Simulation results demonstrate that the proposed algorithm achieves a 100% probability in frame recognition when BER is below 0.2. Even with a BER of 0.25, the recognition probability remains above 90%, which exhibits a performance improvement of over 60% compared with traditional algorithms. This work addresses the shortcomings of existing methods under high error conditions, and the idea of converting sequences into RGB images also provides a reliable solution for frame synchronization in challenging communication environments.

PMID:39460248 | DOI:10.3390/s24206767

Sensors (Basel). 2024 Oct 21;24(20):6754. doi: 10.3390/s24206754.

ABSTRACT

Spatial resolution enhancement in remote sensing data aims to augment the level of detail and accuracy in images captured by satellite sensors. We proposed a novel spatial resolution enhancement framework using the convolutional attention-based token mixer method. This approach leveraged spatial context and semantic information to improve the spatial resolution of images. This method used the multi-head convolutional attention block and sub-pixel convolution to extract spatial and spectral information and fused them using the same technique. The multi-head convolutional attention block can effectively utilize the local information of spatial and spectral dimensions. The method was tested on two kinds of data types, which were the visual-thermal dataset and the visual-hyperspectral dataset. Our method was also compared with the state-of-the-art methods, including traditional methods and deep learning methods. The experiment results showed that the method was effective and outperformed state-of-the-art methods in overall, spatial, and spectral accuracies.

PMID:39460237 | DOI:10.3390/s24206754

Sensors (Basel). 2024 Oct 21;24(20):6757. doi: 10.3390/s24206757.

ABSTRACT

Internal disease in asphalt pavement is a crucial indicator of pavement health and serves as a vital basis for maintenance and rehabilitation decisions. It is closely related to the optimization and allocation of funds by highway maintenance management departments. Accurate and rapid identification of internal pavement diseases is essential for improving overall pavement quality. This study aimed to identify internal pavement diseases using deep learning algorithms, thereby improving the efficiency of determining internal pavement diseases. In this work, a multi-view recognition algorithm model based on deep learning is proposed, with attention fusion mechanisms embedded both between channels and between views. By comparing and analyzing the training and recognition results of different neural networks, it was found that the multi-view recognition algorithm model based on attention fusion demonstrates the best performance in identifying internal pavement diseases.

PMID:39460235 | DOI:10.3390/s24206757

Sensors (Basel). 2024 Oct 21;24(20):6750. doi: 10.3390/s24206750.

ABSTRACT

Pneumonia is a form of acute respiratory infection affecting the lungs. Symptoms of viral and bacterial pneumonia are similar. Rapid diagnosis of the disease is difficult, since polymerase chain reaction-based methods, which have the greatest reliability, provide results in a few hours, while ensuring high requirements for compliance with the analysis technology and professionalism of the personnel. This study proposed a Concatenated CNN model for pneumonia detection combined with a fuzzy logic-based image improvement method. The fuzzy logic-based image enhancement process is based on a new fuzzification refinement algorithm, with significantly improved image quality and feature extraction for the CCNN model. Four datasets, original and upgraded images utilizing fuzzy entropy, standard deviation, and histogram equalization, were utilized to train the algorithm. The CCNN's performance was demonstrated to be significantly improved by the upgraded datasets, with the fuzzy entropy-added dataset producing the best results. The suggested CCNN attained remarkable classification metrics, including 98.9% accuracy, 99.3% precision, 99.8% F1-score, and 99.6% recall. Experimental comparisons showed that the fuzzy logic-based enhancement worked significantly better than traditional image enhancement methods, resulting in higher diagnostic precision. This study demonstrates how well deep learning models and sophisticated image enhancement techniques work together to analyze medical images.

PMID:39460230 | DOI:10.3390/s24206750

Sensors (Basel). 2024 Oct 20;24(20):6747. doi: 10.3390/s24206747.

ABSTRACT

Bolt loosening detection is crucial for ensuring the safe operation of equipment. This paper presents a vision-based real-time detection method that identifies bolt loosening by recognizing anti-loosening line markers at bolt connections. The method employs the YOLOv10-S deep learning model for high-precision, real-time bolt detection, followed by a two-step Fast-SCNN image segmentation technique. This approach effectively isolates the bolt and nut regions, enabling accurate extraction of the anti-loosening line markers. Key intersection points are calculated using ellipse and line fitting techniques, and the loosening angle is determined through spatial projection transformation. The experimental results demonstrate that, for high-resolution images of 2048 × 1024 pixels, the proposed method achieves an average angle detection error of 1.145° with a detection speed of 32 FPS. Compared to traditional methods and other vision-based approaches, this method offers non-contact measurement, real-time detection capabilities, reduced detection error, and general adaptability to various bolt types and configurations, indicating significant application potential.

PMID:39460227 | DOI:10.3390/s24206747

Sensors (Basel). 2024 Oct 20;24(20):6746. doi: 10.3390/s24206746.

ABSTRACT

Due to the small size of vehicle targets, complex background environments, and the discrete scattering characteristics of high-resolution synthetic aperture radar (SAR) images, existing deep learning networks face challenges in extracting high-quality vehicle features from SAR images, which impacts vehicle localization accuracy. To address this issue, this paper proposes a vehicle localization method for SAR images based on feature reconstruction and aggregation with rotating boxes. Specifically, our method first employs a backbone network that integrates the space-channel reconfiguration module (SCRM), which contains spatial and channel attention mechanisms specifically designed for SAR images to extract features. The network then connects a progressive cross-fusion mechanism (PCFM) that effectively combines multi-view features from different feature layers, enhancing the information content of feature maps and improving feature representation quality. Finally, these features containing a large receptive field region and enhanced rich contextual information are input into a rotating box vehicle detection head, which effectively reduces false alarms and missed detections. Experiments on a complex scene SAR image vehicle dataset demonstrate that the proposed method significantly improves vehicle localization accuracy. Our method achieves state-of-the-art performance, which demonstrates the superiority and effectiveness of the proposed method.

PMID:39460226 | DOI:10.3390/s24206746

Sensors (Basel). 2024 Oct 20;24(20):6740. doi: 10.3390/s24206740.

ABSTRACT

With the rapid development of wearable biosensor technology, the combination of head-mounted displays and augmented reality (AR) technology has shown great potential for health monitoring and biomedical diagnosis applications. However, further optimizing its performance and improving data interaction accuracy remain crucial issues that must be addressed. In this study, we develop smart glasses based on augmented reality and eye tracking technology. Through real-time information interaction with the server, the smart glasses realize accurate scene perception and analysis of the user's intention and combine with mixed-reality display technology to provide dynamic and real-time intelligent interaction services. A multi-level hardware architecture and optimized data processing process are adopted during the research process to enhance the system's real-time accuracy. Meanwhile, combining the deep learning method with the geometric model significantly improves the system's ability to perceive user behavior and environmental information in complex environments. The experimental results show that when the distance between the subject and the display is 1 m, the eye tracking accuracy of the smart glasses can reach 1.0° with an error of no more than ±0.1°. This study demonstrates that the effective integration of AR and eye tracking technology dramatically improves the functional performance of smart glasses in multiple scenarios. Future research will further optimize smart glasses' algorithms and hardware performance, enhance their application potential in daily health monitoring and medical diagnosis, and provide more possibilities for the innovative development of wearable devices in medical and health management.

PMID:39460220 | DOI:10.3390/s24206740

Sensors (Basel). 2024 Oct 20;24(20):6737. doi: 10.3390/s24206737.

ABSTRACT

The rapid adoption of Unmanned Aerial Vehicles (UAVs) in the construction industry has revolutionized safety, surveying, quality monitoring, and maintenance assessment. UAVs are increasingly used to prevent accidents caused by falls from heights or being struck by falling objects by ensuring workers comply with safety protocols. This study focuses on leveraging UAV technology to enhance labor safety by monitoring the use of personal protective equipment, particularly helmets, among construction workers. The developed UAV system utilizes the tensorflow technique and an alert system to detect and identify workers not wearing helmets. Employing the high-precision, high-speed, and widely applicable Faster R-CNN method, the UAV can accurately detect construction workers with and without helmets in real-time across various site conditions. This proactive approach ensures immediate feedback and intervention, significantly reducing the risk of injuries and fatalities. Additionally, the implementation of UAVs minimizes the workload of site supervisors by automating safety inspections and monitoring, allowing for more efficient and continuous oversight. The experimental results indicate that the UAV system's high precision, recall, and processing capabilities make it a reliable and cost-effective solution for improving construction site safety. The precision, mAP, and FPS of the developed system with the R-CNN are 93.1%, 58.45%, and 27 FPS. This study demonstrates the potential of UAV technology to enhance safety compliance, protect workers, and improve the overall quality of safety management in the construction industry.

PMID:39460217 | DOI:10.3390/s24206737

J Med Internet Res. 2024 Oct 25;26:e57750. doi: 10.2196/57750.

ABSTRACT

BACKGROUND: The rapid growth of research in artificial intelligence (AI) and machine learning (ML) continues. However, it is unclear whether this growth reflects an increase in desirable study attributes or merely perpetuates the same issues previously raised in the literature.

OBJECTIVE: This study aims to evaluate temporal trends in AI/ML studies over time and identify variations that are not apparent from aggregated totals at a single point in time.

METHODS: We identified AI/ML studies registered on ClinicalTrials.gov with start dates between January 1, 2010, and December 31, 2023. Studies were included if AI/ML-specific terms appeared in the official title, detailed description, brief summary, intervention, primary outcome, or sponsors' keywords. Studies registered as systematic reviews and meta-analyses were excluded. We reported trends in AI/ML studies over time, along with study characteristics that were fast-growing and those that remained unchanged during 2010-2023.

RESULTS: Of 3106 AI/ML studies, only 7.6% (n=235) were regulated by the US Food and Drug Administration. The most common study characteristics were randomized (56.2%; 670/1193; interventional) and prospective (58.9%; 1126/1913; observational) designs; a focus on diagnosis (28.2%; 335/1190) and treatment (24.4%; 290/1190); hospital/clinic (44.2%; 1373/3106) or academic (28%; 869/3106) sponsorship; and neoplasm (12.9%; 420/3245), nervous system (12.2%; 395/3245), cardiovascular (11.1%; 356/3245) or pathological conditions (10%; 325/3245; multiple counts per study possible). Enrollment data were skewed to the right: maximum 13,977,257; mean 16,962 (SD 288,155); median 255 (IQR 80-1000). The most common size category was 101-1000 (44.8%; 1372/3061; excluding withdrawn or missing), but large studies (n>1000) represented 24.1% (738/3061) of all studies: 29% (551/1898) of observational studies and 16.1% (187/1163) of trials. Study locations were predominantly in high-income countries (75.3%; 2340/3106), followed by upper-middle-income (21.7%; 675/3106), lower-middle-income (2.8%; 88/3106), and low-income countries (0.1%; 3/3106). The fastest-growing characteristics over time were high-income countries (location); Europe, Asia, and North America (location); diagnosis and treatment (primary purpose); hospital/clinic and academia (lead sponsor); randomized and prospective designs; and the 1-100 and 101-1000 size categories. Only 5.6% (47/842) of completed studies had results available on ClinicalTrials.gov, and this pattern persisted. Over time, there was an increase in not only the number of newly initiated studies, but also the number of completed studies without posted results.

CONCLUSIONS: Much of the rapid growth in AI/ML studies comes from high-income countries in high-resource settings, albeit with a modest increase in upper-middle-income countries (mostly China). Lower-middle-income or low-income countries remain poorly represented. The increase in randomized or prospective designs, along with 738 large studies (n>1000), mostly ongoing, may indicate that enough studies are shifting from an in silico evaluation stage toward a prospective comparative evaluation stage. However, the ongoing limited availability of basic results on ClinicalTrials.gov contrasts with this field's rapid advancements and the public registry's role in reducing publication and outcome reporting biases.

PMID:39454187 | DOI:10.2196/57750

PLoS One. 2024 Oct 25;19(10):e0310002. doi: 10.1371/journal.pone.0310002. eCollection 2024.

ABSTRACT

Egypt faces extreme traffic congestion in its cities, which results in long travel times, large lines of parked cars, and increased safety hazards. Our study suggests a multi-modal approach that combines critical infrastructure improvements with cutting-edge technologies to address the ubiquitous problem of traffic congestion. Assuring vehicles owners of their timely arrival, cutting down on fuel usage, and improving communication using deep learning approach and optimization algorithm within the potential of IoT enabled 5G framework are the main goals. The traffic management system incorporates detection cameras, Raspberry Pi 3 microcontroller, an Android application, cloud connectivity, and traditional traffic lights that are powered using PV modules and batteries to secure the traffic controllers operation in case of grid outage and assure service continuity. The model examines the difficulties associated with Internet of Things (IoT) communication, highlighting possible interference from device-to-device (D2D) devices and cellular user equipment. This all-encompassing strategy aims to reduce fuel consumption, increase road safety and improve traffic efficiency. The model predicts a significant increase in Egypt's urban mobility by utilizing the possibilities of IoT and 5G technologies, which would improve Egypt's towns' livability and efficiency. The goal of this paper is to modernize Egypt's traffic management system and bring it into compliance with global guidelines for intelligent transportation networks.

PMID:39453902 | DOI:10.1371/journal.pone.0310002

IEEE Trans Image Process. 2024 Oct 25;PP. doi: 10.1109/TIP.2024.3482191. Online ahead of print.

ABSTRACT

Recent advances in deep learning algorithms have shown impressive progress in image copy-move forgery detection (CMFD). However, these algorithms lack generalizability in practical scenarios where the copied regions are not present in the training images, or the cloned regions are part of the background. Additionally, these algorithms utilize convolution operations to distinguish source and target regions, leading to unsatisfactory results when the target regions blend well with the background. To address these limitations, this study proposes a novel end-to-end CMFD framework that integrates the strengths of conventional and deep learning methods. Specifically, the study develops a deep cross-scale PatchMatch (PM) method that is customized for CMFD to locate copy-move regions. Unlike existing deep models, our approach utilizes features extracted from high-resolution scales to seek explicit and reliable point-to-point matching between source and target regions. Furthermore, we propose a novel pairwise rank learning framework to separate source and target regions. By leveraging the strong prior of point-to-point matches, the framework can identify subtle differences and effectively discriminate between source and target regions, even when the target regions blend well with the background. Our framework is fully differentiable and can be trained end-to-end. Comprehensive experimental results highlight the remarkable generalizability of our scheme across various copy-move scenarios, significantly outperforming existing methods.

PMID:39453802 | DOI:10.1109/TIP.2024.3482191

IEEE Trans Image Process. 2024 Oct 25;PP. doi: 10.1109/TIP.2024.3484173. Online ahead of print.

ABSTRACT

High dynamic range (HDR) video offers a more realistic visual experience than standard dynamic range (SDR) video, while introducing new challenges to both compression and transmission. Rate control is an effective technology to overcome these challenges, and ensure optimal HDR video delivery. However, the rate control algorithm in the latest video coding standard, versatile video coding (VVC), is tailored to SDR videos, and does not produce well coding results when encoding HDR videos. To address this problem, a data-driven λ-domain rate control algorithm is proposed for VVC HDR intra frames in this paper. First, the coding characteristics of HDR intra coding are analyzed, and a piecewise R-λ model is proposed to accurately determine the correlation between the rate (R) and the Lagrange parameter λ for HDR intra frames. Then, to optimize bit allocation at the coding tree unit (CTU)-level, a wavelet-based residual neural network (WRNN) is developed to accurately predict the parameters of the piecewise R-λ model for each CTU. Third, a large-scale HDR dataset is established for training WRNN, which facilitates the applications of deep learning in HDR intra coding. Extensive experimental results show that our proposed HDR intra frame rate control algorithm achieves superior coding results than the state-of-the-art algorithms. The source code of this work will be released at https://github.com/TJU-Videocoding/WRNN.git.

PMID:39453801 | DOI:10.1109/TIP.2024.3484173

IEEE Trans Neural Syst Rehabil Eng. 2024 Oct 25;PP. doi: 10.1109/TNSRE.2024.3486444. Online ahead of print.

ABSTRACT

The objective of this study was to propose a novel strategy for detecting upper-limb motion intentions from mechanical sensor signals using deep and heterogeneous transfer learning techniques. Three sensor types, surface electromyography (sEMG), force-sensitive resistors (FSRs), and inertial measurement units (IMUs), were combined to capture biometric signals during arm-up, hold, and arm-down movements. To distinguish motion intentions, deep learning models were constructed using the CIFAR-ResNet18 and CIFAR-MobileNetV2 architectures. The input features of the source models were sEMG, FSR, and IMU signals. The target model was trained using only FSR and IMU sensor signals. Optimization techniques determined appropriate layer structures and learning rates of each layer for effective transfer learning. The source model on CIFAR-ResNet18 exhibited the highest performance, achieving an accuracy of 95% and an F-1 score of 0.95. The target model with optimization strategies performed comparably to the source model, achieving an accuracy of 93% and an F-1 score of 0.93. The results show that mechanical sensors alone can achieve performance comparable to models including sEMG. The proposed approach can serve as a convenient and precise algorithm for human-robot collaboration in rehabilitation assistant robots.

PMID:39453796 | DOI:10.1109/TNSRE.2024.3486444

IEEE Rev Biomed Eng. 2024 Oct 25;PP. doi: 10.1109/RBME.2024.3486439. Online ahead of print.

ABSTRACT

Cardiac digital twins (CDTs) are personalized virtual representations used to understand complex cardiac mechanisms. A critical component of CDT development is solving the ECG inverse problem, which enables the reconstruction of cardiac sources and the estimation of patient-specific electrophysiology (EP) parameters from surface ECG data. Despite challenges from complex cardiac anatomy, noisy ECG data, and the ill-posed nature of the inverse problem, recent advances in computational methods have greatly improved the accuracy and efficiency of ECG inverse inference, strengthening the fidelity of CDTs. This paper aims to provide a comprehensive review of the methods of solving ECG inverse problem, the validation strategies, the clinical applications, and future perspectives. For the methodologies, we broadly classify state-of-the-art approaches into two categories: deterministic and probabilistic methods, including both conventional and deep learning-based techniques. Integrating physics laws with deep learning models holds promise, but challenges such as capturing dynamic electrophysiology accurately, accessing accurate domain knowledge, and quantifying prediction uncertainty persist. Integrating models into clinical workflows while ensuring interpretability and usability for healthcare professionals is essential. Overcoming these challenges will drive further research in CDTs.

PMID:39453795 | DOI:10.1109/RBME.2024.3486439

IEEE Trans Biomed Eng. 2024 Nov;71(11):3160-3169. doi: 10.1109/TBME.2024.3409642.

ABSTRACT

Necrotizing Enterocolitis (NEC) is a devastating condition affecting prematurely born neonates. Reviewing Abdominal X-rays (AXRs) is a key step in NEC diagnosis, staging and treatment decision-making, but poses significant challenges due to the subtle, difficult-to-identify radiological signs of the disease. In this paper, we propose AIDNEC - AI Diagnosis of NECrotizing enterocolitis, a deep learning method to automatically detect and stratify the severity (surgical or medical) of NEC from no pathology in AXRs. The model is trainable end-to-end and integrates a Detection Transformer and Graph Convolution modules for localizing discriminative areas in AXRs, used to formulate subtle local embeddings. These are then combined with global image features to perform Fine-Grained Visual Classification (FGVC). We evaluate AIDNEC on our GOSH NEC dataset of 1153 images from 334 patients, achieving 79.7% accuracy in classifying NEC against No Pathology. AIDNEC outperforms the backbone by 2.6%, FGVC models by 2.5% and CheXNet by 4.2%, with statistically significant (two-tailed p 0.05) improvements, while providing meaningful discriminative regions to support the classification decision. Additional validation in the publicly available Chest X-ray14 dataset yields comparable performance to state-of-the-art methods, illustrating AIDNEC's robustness in a different X-ray classification task.

PMID:39453790 | DOI:10.1109/TBME.2024.3409642

ACS Appl Mater Interfaces. 2024 Oct 25. doi: 10.1021/acsami.4c11997. Online ahead of print.

ABSTRACT

Doping plays a crucial role in determining the critical temperature (Tc) of superconductors, yet accurately predicting its effects remains a significant challenge. Here, we introduce a novel doping descriptor that captures the complex influence of dopants on superconductivity. By integrating the doping descriptor with elemental and physical features within a Mixture of Experts (MoE) model, we achieve a remarkable R2 of 0.962 for Tc prediction, surpassing all published prediction models. Our approach successfully identifies optimal doping levels in the Bi2-xPbxSr2Ca2-yCuyOz system, with predictions closely aligning with experimental results. Leveraging this model, we screen compounds from the Inorganic Crystal Structure Database and employ a generative approach to explore new doped superconductors. This process reveals 40 promising candidates for high Tc superconductivity among existing and hypothetical doped materials. By explicitly accounting for doping effects, our method offers a powerful tool for guiding the experimental discovery of new superconductors, potentially accelerating progress in high-temperature superconductivity research and opening new avenues for material design.

PMID:39453724 | DOI:10.1021/acsami.4c11997

Med Biol Eng Comput. 2024 Oct 25. doi: 10.1007/s11517-024-03219-4. Online ahead of print.

ABSTRACT

Cerebrovascular image segmentation is one of the crucial tasks in the field of biomedical image processing. Due to the variable morphology of cerebral blood vessels, the traditional convolutional kernel is weak in perceiving the structure of elongated blood vessels in the brain, and it is easy to lose the feature information of the elongated blood vessels during the network training process. In this paper, a vascular convolutional U-network (VCU-Net) is proposed to address these problems. This network utilizes a new convolution (vascular convolution) instead of the traditional convolution kernel, to extract features of elongated blood vessels in the brain with different morphologies and orientations by adaptive convolution. In the network encoding stage, a new feature splicing method is used to combine the feature tensor obtained through vascular convolution with the original tensor to provide richer feature information. Experiments show that the DSC and IOU of the proposed method are 53.57% and 69.74%, which are improved by 2.11% and 2.01% over the best performance of the GVC-Net among several typical models. In image visualization, the proposed network has better segmentation performance for complex cerebrovascular structures, especially in dealing with elongated blood vessels in the brain, which shows better integrity and continuity.

PMID:39453556 | DOI:10.1007/s11517-024-03219-4

JACC Heart Fail. 2024 Oct 9:S2213-1779(24)00618-8. doi: 10.1016/j.jchf.2024.08.012. Online ahead of print.

ABSTRACT

BACKGROUND: The lack of automated tools for measuring care quality limits the implementation of a national program to assess guideline-directed care in heart failure with reduced ejection fraction (HFrEF).

OBJECTIVES: The authors aimed to automate the identification of patients with HFrEF at hospital discharge, an opportunity to evaluate and improve the quality of care.

METHODS: The authors developed a novel deep-learning language model for identifying patients with HFrEF from discharge summaries of hospitalizations with heart failure at Yale New Haven Hospital during 2015 to 2019. HFrEF was defined by left ventricular ejection fraction <40% on antecedent echocardiography. The authors externally validated the model at Northwestern Medicine, community hospitals of Yale, and the MIMIC-III (Medical Information Mart for Intensive Care III) database.

RESULTS: A total of 13,251 notes from 5,392 unique individuals (age 73 ± 14 years, 48% women), including 2,487 patients with HFrEF (46.1%), were used for model development (train/held-out: 70%/30%). The model achieved an area under receiver-operating characteristic curve (AUROC) of 0.97 and area under precision recall curve (AUPRC) of 0.97 in detecting HFrEF on the held-out set. The model had high performance in identifying HFrEF with AUROC = 0.94 and AUPRC = 0.91 on 19,242 notes from Northwestern Medicine, AUROC = 0.95 and AUPRC = 0.96 on 139 manually abstracted notes from Yale community hospitals, and AUROC = 0.91 and AUPRC = 0.92 on 146 manually reviewed notes from MIMIC-III. Model-based predictions of HFrEF corresponded to a net reclassification improvement of 60.2 ± 1.9% compared with diagnosis codes (P < 0.001).

CONCLUSIONS: The authors developed a language model that identifies HFrEF from clinical notes with high precision and accuracy, representing a key element in automating quality assessment for individuals with HFrEF.

PMID:39453355 | DOI:10.1016/j.jchf.2024.08.012